Skip to main content

Keep Your AI Grounded in Reality

Grounded Docs MCP Server eliminates AI hallucinations by providing always-current, version-specific documentation to your AI coding assistant.

Why Grounded Docs?

The open-source alternative to Context7, Nia, and Ref.Tools.

Up-to-Date Context

Fetches documentation directly from official sources on demand, ensuring your AI always has current information.

Version-Specific

Queries target the exact library versions in your project, eliminating version mismatch issues.

Reduces Hallucinations

Grounds LLMs in real documentation from official sources, dramatically reducing false information.

Private & Local

Runs entirely on your machine. Your code never leaves your network.

Quick Start

1

Start the server

Launch the MCP server with npx (requires Node.js 22+):
npx @arabold/docs-mcp-server@latest
2

Open the Web UI

Navigate to http://localhost:6280 to add documentation sources.
3

Connect your AI client

Add this to your MCP settings (e.g., claude_desktop_config.json):
{
  "mcpServers": {
    "docs-mcp-server": {
      "type": "sse",
      "url": "http://localhost:6280/sse"
    }
  }
}

Full Installation Guide

See complete setup instructions for Docker, Node.js, and embedded mode.

Key Features

Multiple Sources

Index websites, GitHub repos, npm/PyPI packages, and local files.

Semantic Search

Vector embeddings with OpenAI, Gemini, Azure, AWS, and Ollama.

Rich File Support

HTML, Markdown, PDF, Word, Excel, PowerPoint, and source code.

MCP Compatible

Works with Claude, Cline, Roo, and any MCP-compatible client.

Distributed Mode

Scale processing across multiple workers.

CLI Tool

Command-line interface for scripting and automation.

Get Started

Quickstart Guide

Get up and running in under 5 minutes.

Configuration

Configure embedding models and advanced options.

CLI Reference

Learn about all available CLI commands.

MCP Tools

Explore the MCP protocol tools available.

Architecture

Learn how Grounded Docs works under the hood.

System Overview

Understand the core architecture and components.

Deployment Modes

Choose between unified or distributed deployment.

Build docs developers (and LLMs) love